Martingales

The fundamental property of a Martingale is that its expected value remains constant over time, regardless of the past history of the process.

The Martingale strategy in a fair coin toss involves doubling the bet after each loss in an attempt to recoup the losses and eventually make a profit.

The expected value of a fair Martingale is always zero, regardless of the initial bet. This is because the gains and losses over time balance out.

A Martingale is a special type of random walk where the expected value remains constant over time.

Geometric Brownian motion is an example of a continuous-time Martingale, commonly used in financial modeling.

Doob's Martingale Convergence Theorem provides sufficient conditions for the convergence of Martingales, ensuring that they have well-defined limits.

The gambler's ruin problem addresses the probability that a gambler with a finite amount of money will eventually lose everything.

Martingales can be represented as conditional expectations of the future values of a stochastic process given the past information.

Martingales are widely used in finance for pricing options, developing hedging strategies, and managing risk.

The optional stopping theorem provides conditions under which a Martingale can be stopped at a random time without affecting its Martingale property.

Supermartingales are a subclass of Martingales where the expected value of the future values is less than or equal to the current value.

The Poisson process is an example of a discrete-time supermartingale, commonly used in modeling the arrival of events over time.

Submartingales are a subclass of Martingales where the expected value of the future values is greater than or equal to the current value.

Geometric Brownian motion can be a submartingale or a supermartingale, depending on the drift and volatility parameters.

The square-integrable condition allows for the application of powerful mathematical techniques, such as stochastic calculus, to study the properties of Martingales.